University of the Pacific Scholarly Commons

University of the Pacific Theses and Dissertations Graduate School

1961

An experimental investigation of coloring matter in Indica flowers

James Clinton Vogt University of the Pacific

Follow this and additional works at: https://scholarlycommons.pacific.edu/uop_etds

Part of the Chemistry Commons

Recommended Citation Vogt, James Clinton. (1961). An experimental investigation of coloring matter in flowers. University of the Pacific, Thesis. https://scholarlycommons.pacific.edu/uop_etds/368

This Thesis is brought to you for free and open access by the Graduate School at Scholarly Commons. It has been accepted for inclusion in University of the Pacific Theses and Dissertations by an authorized administrator of Scholarly Commons. For more information, please contact [email protected]. AN EXPERIMIDTTAL IliVEf)TIGATION OF COLORING MATTER It IN CANNA INDICA FLOWERS

A Thesis Submitted to the Faculty of the University of the Pacific

In Partial Fulfillment for the Degzoee of Master of .Arts in the Department of Chemistry

by James Clinton Vogt !.tt June 1961 ACKNOWLEDGMENT

The writer wishes to express his appreciation to the members o:f: the faculty of the Chemistry .Department of' the University of' the Pacific for their assistance during the period of study and. especi.ally to express his gratitude to Dr. Emerson G.- Cobb for his guidance and criticisms in this work .. TABLE ·OF OOWTENTS Page

INTRODUCTION 1 · Coloring matter in f1owe:rs 1 Basic structures 1 Anthoeyanidin fundamental tnes 8 Chemistry o.f a.nthocyanins, anthoeyanidins · 11 Degradation products of'. 13 Synthesis of arithocyanins and anthoeyanidins 14 Proper tie a of the 17 Absorption spectra 20 Oeeurrenee 21 Isols.tion 21

EXPERir~TAL RESEARCH 22 lJfethanol extraction 22 So.lubili ty tests 23 Other extraction and tests 2.3 Degradation products 25 Figure 1 - transmission spectrum 27 SUMMARY 28 BIBLIOGRAPHY 29 INTRODUCTION

The chemist has long been attracted to the investi­ gation of natural and artificial coloring matters for a variety o1' reasons, including not only color-pleasure,

the industrial importance o1' dyestuf';fs end pigments 1 but also on account of the fact that visible color more than any other property raeilitates the experimental study of ·organic substances whether by analysis or by s,nthesis. Color .furniShes a standard or homogeneity or a measure ef concentration, is an invaJ.ua.ble guide in the search .for methods of separation and purtifieation, and it at once indicates by its appearance or disappearance the occur• renee of a 'che':mieal reaction. Al.though the chemical nature of this coloring matter vas not known by early man, the search for this knowledge was started ma:n:y years ago. As long ago as

16641 Robert Boyle published "Experiments and Considera­ tions Touching Colours," in which he examined results when extracts from. :f'lowe::rs were treated with acids and alkalies. In eases whez-e pigments could be utilized for dyeing. the practical value became so great that as soon as a structural theol"y of organic chemistry had been 2 ..

developed, chemical constituents of pigments ware determined and the materials synthesized., This was particularly the ease where production of natural substances could not keep --up with the del.tlSnd. The- research was carried out for the purpose of isol:ating and identifying the eol~:Ping matter in flowers of the canna indica (canna. lily) varia ty. It was done. by

extract-ing the' pigment .from flower with methanol, amyl alcohol, and other solvents, and e:rystallizing the pigment a.s a salt. Physical properties such as solubility, color I"eaetions, spectra and othei"s were examined for this pigment. Chemical properties and degradation products were also examined in order to identify the chemical structure of the pigment.

Two important groups of pigments that account for most red, blue and yellow colors in flowers are know-n as anthoeyanins and anthoxanthins.. In general the darker pig­ ments are due to the group and the lighter pig­ ments to the anthoxanthin group. IVIembers of both the an­ thocyanin and the anthoxanthin groups of pigments have a heterocyclic structure al'ld a large number have been iso... la.ted and identified. Red and blue water soluble pigments belong to the anthocyanin group while yellow water soluble pigments be ... long to the anthoxanthin group which eontains .flavone or fla.vonol compounds (20}. The anthoxanthin pigments eonsti• tute the largest and most wi.dely distributed group in na.... ture • and are often a.sso ciated with other pigments and tan- nine. A second classification of plant pigments is the

. ' '' ' . plastid group" the members of which are associated with

the protoplasmic stl'Ueture of the plant,. and a second group which generally exists in solution in the cell sap and are called snthocyanins. The plastid pigments are not soluble in organic solvents such as ether and benzene. Included in this gl!'oup are , zanthoplylls and. .

~ey differ considerably from the water soluble pigments. The anthoeyanins belong to a larger group o.f glyeosides. These anthocyanin pigments account :for shades of blue, pur­ ple,. viola t 1 mauve, magenta~ Slld nea:rly all the reds in

.flowers, rru1 ts6 and plant stems. They are always . c.ombine.d with sugars, are water soluble and usually dis-

so 1 ved. in the e ell sap • They may, however, exi st in &1 ther

an amorphous or a c:rystalline state .. ($} Sugar :free pigments or aglucons are called anthoey... anidins. .Anthoeyanins were extracted early from. flowers end leaves with amyl alcohol, then converted to the corres­ ponding anthoeyanidins on the addition of acids such as hydrochloric acid. St. Jonekco (21) concluded that the presence of antho­ eyan.ins) anthoeyanidins, and pseudo bases together indicates their close interrelationship and their relation to the dis­ appearance of' the red color in . Anthocyanidins appear characteristic of pure· red oi"gans' and do not exist as colored pigme'nts in a. .free condition. in sll colored tissue containing . . anthocyanin, It ls further conclude.d that canna, among other flowers having purple~red, purple, violet or blue colorations do ·not contain anthoeyanidins but contain an anthocyanin pig­ ment, the coloration o.f ltl.ioh varies in each !:peeies o.f plant, and an intense pure yellow pigment which dissolves . . . in amyl alcohol like the anthoeyanidins but is not a pseudo base. Robinson (10) concluded that an thoeyanins and osn be regar>ded as being produced divergently from two plant substances. One is present in a limited amount and is a component of all pigments while the second is produced in an amount and variety depend;~nt on fa.cto'%-ial in.fl uenees a.nd ·... · interactions. Principal factor-s a:f':f"eeting colors of anthocyanin and anthoxanthin pigments in nature are {1) nature and con­ centration of anthocyanin, anthoxanthin pigments and other colored 'substances pr-esent; (2) the state of anthocyanin and anthoxanthin pigments in solution, which is determined ,;;;;;;;;;;--~~-~~------"--'--

5.

by the p~ o~ the cell sap and the presence or absence of protective colloids of the polysaccharide group; and

{3) the presence or absence of other p~gments such as tan• nins., also the effect of alkaloids and traces of iron and other metals that form complex combinations. i'iith regard

to variable colors of flower petals, Robinson (11) stated that the factor is the concentration of anthocyanin and the ratio of this to the concentration of co•pigmentsof the tannin and :flavonol cla.s.ses. Pale contains much less anthocyanin than darker colors end with higher concentra­

tion -the po,... -pigment is unable to mod.ify the color of all

the- antho~yanin.- The character of the anthoeyanidin

resor;tating system is seriously_si'f'ected by the presence

of mineral salts and by the pB of the environment. Conse­ quently colors of pigments of flowers sometimes vary

ma.I'kedly with type of soil. . Thus eye.nin1 the pigment of

red ros.e and blue cornflower, is pale violet in neutral solution, red in dilute acid and blue in dilute base, through which color reactions ·it acts as an indicator. The color of .flowers_, however, should not be taken as a reliable measure of the environmental pR.

Cram and Hammond {2) state that anthoeyanins a.:roe one of the main classes of plant pigments. They occur in .flowers and f'ruits as glycosides, hydrolysis of which pro­ vides colored aglycones known as anthocyanidins. Vivid blues and reds of anthocyanins are associated with the . 6 •. distribution o:r ·positive chal"'ge throughout an aryl-substi ... tuted·ch.roman ring system (I}• I.

>

< > < > etc.

An:thoeyanidins are usually isola ted in the form of chloride salts, are frequently hydroxylated in some or all of the 3, 5, 1, 3t, 4', 5' positions. Using this explana­ tion, stPUetural formulas can be shown for the red cyanin cation {II), the violet cyanin color base (III), and the blue cyanin anion (IV). Cyanins in genera.l differ only in the .number and position of hydroxyl groups and in the

character of sugars to which they are attaehed.- II. III.

HO =o

O.glucoside .glucoside "'"jiiiiii__.--~--H-·-----~---

IV.

OH )

0-glucoside The b.a.sic unit structure of' anthoxanthin is benzo- pyrene {V) and that of anthocyanin is benzopyr.ylium (VI}, both of which contain the basic quadrivalant oxygen atom. v. VI.

This quadrivalent oxygen atom can form additive compounds

v;i th acids producing oxonium sel ts.. These oxonium salts

of anthoxanth:tns are unstable, are usually precipitated by using lead acetate. Oxonium salts of anthocyanins on the

other hand are very stable and frequently occur as such ~

the pl~t. Dilute acids can usually be used to isolate

them from their solutions. Hydrochloric acid" for exa.mple1 can be added to get readily crystallizeable compounds such as 3, 5, 7, trihydroxyflavylium cr.tloride (VII) which is the simplest intact structural unit of ell the anthocyanin salts. a. VII• Cl

..._ ,. 'ii·-- . · The· chloride' salts of anthocys.nins are called anthoeyanidins. The silllplest strudtural :fo:rmu.la of anthoxanthins is :flavone (VIIJ:) and. the basic structure of' the is 3 h:Vdroxyflavone (IX). VIII .. IX ..

H H 0 0

Hydroxy groups may be attached to the 3$ 5~ 71 · 3' • 4', and/or 5' po.si tiona to :form the difi'erent anthoxanthins found. in natui-e. The simplest . anthocyanid.in found in ns.tur.e is· pel­ argcnidin (X). This chloride salt of the anthocyanin :from

which it is derived has hydroxy groups in the 3, 5, 7 ~ 4 1 positions. x.

OH

There are two other fundamental type groups; (.XI), which is 3~ 5, 7~ 3', 4' pentahydroxy 2 phenyl'benzopy:Pilium chloride, and (XII}, t.vhich is 3.- 5., 7 • 3' , 4 t 1 5• he.xahydroxy 2 phenylbenzopyriJ.ium chloride• XII.

HO OH

Three other fundame.ntal type groups are ethe.rs of cyanidin or d.elphinidin. These three are peonidin (XIII) which is

3, 5, 7, 4' tetrahydl:miy _3t methoxy 2 phenylbenzopYI"ilium chloride; malvidin or syringidin {XIV) which is 3, 5, 1, 4' tetrahydroxy 3' 5' dimethoxy 2 phenylbenzopyrilium chloride; and hirsutidin (XV) which is 3, 5, 4i trihydroxy 7, 3', 51 trimethoxy 2 phenylbenzopyr:llium ehleride. 10. XIII. XIV ..

XV.

All six types have been synthesized sq the structure is known. :Robinson and Robinson {17) concluded that almost the whole range ot anthocyanin pigments of flowers, fruits and blossoms is derived from the first three fundamental types of anthoeyanidins l_isted1 namely, pelargonidin, cy'f!in­ idin and delphinidin, by various substitutions in the hy­ droxy groups. They gave as exceptions the bluest antho­ eya.nin found in the beet Celosia eristata~ and Atriplex hortensis as being nitrogenous pigments. At the other end of the seale the most yellow anthocyanin noted in PapaV-er alpinu.-wn and the yellow Iceland Poppy is %-'elated to the flavones. 11. Anthoeyanins are glucosides o'£ a:nthoeyanidins., sug... ars being attached to t.'he .3 or the 3, 5 positions, the .;, 5 diglueosides being the most common. Some of the sugaz:ts are monoglucoside, monogalactieide;. rhamnoside o%' pento- side.

All members yield• when boiled with a dilute mine~al acid such as hydrochlol"ie, an. anthocyan! din and a sugar

ol" severaal sugus, Gilman (5). Some of the membe:rs, forr example delphinin, yield in addition to the

and a sugar or sag~s, a third eo:mponent which is .invariably

an organic acid. .Pelargonin (XVI), the 3 1 5 diglucoside ot pela:rgonidin is typical of the .), 5 diglueosides, the m.ost widely distributed group of anthoc;yanins .. XVI. Cl _.A.. 0 HO OH

The acid radicals can be either in ester oombina.Uon with one of the hydroxyl groups of the pigment nucleus or at- .. taehed to an hydl"o.xyl of the sugar component.

Wills tatter aild eo-workers (2.3 1 24, 25, 26) estab­ lished the main features of the chemistry of anthoeyanins 12 .. which were recognized ~s saeoharides, occasionally aey­ lated. of the anthoeyanidi:ns. They f'ound that the antho-. eyanins exhibit sm.phote:ric ehai"aeter, fo~ing salts with both acids and bases. Thus the violet pig:.m.$nt eyanin, · ·· which ea.n be isolated. from blue e.orntle>wers, red roses. deep red dahlias and other flowe:rs forms a blue sodium salt and a red hydrochloride.. Hydrolysis of the latter by hot aqueous hydrochloric acid into C7anidin chloride· and glucose can.· be shown by the :following equation.

cyanilJ.. cyanidin glucose chloride chlo:r>id.e Other anthocyanin chlorides can be hydrolized with water to give the anthocyanidin chloride and the sugar that was attached. The oeeurrenee of the 2 phenyl.benzopyrilium. nucleus (XVII} in various anthocyanins was origilU1lly established by Willstatter {2.3) through an alkaline fUsion of sugar free pigments. Empirical formulas of the three parent XVII.

classes of anthocyanidins show that they differ from each other by a single oxygen atom. :Pelargonidin ehloride C:t!)Hll0_$01 or O:t9f70Cl (OR)4 Cyanidin chloride C1#1106Cl or Gl,$H60Cl (OH)S

Delphinidin ehlori.de C15H11~01 or' 015li.5001 {OH}6 These anthocyailidi:ns, which are .free of' sugars,· degrade upon fusion with potassium hydroxide into two simple· prod• ucts, one of' Which is •a phenol, and the o tbe-r a phenol- . ()arboxylie acid. The phenol. is phloroglucinol (XVIII) in all three reactions., end the pheno1carboqlio aeid is in the orde.r named above, respectively; p•hydro.x.ybenzoie acid• (XIX) protocatechuic acid. (XX) and gallie acid (XXI}. HO XVIII. XIX. XXI. OG.u OOOH.,, OOOH'-' ~:9.9 ... ,.· ... OH HOC . . ·. HOG.. .OH ~ . OH ~ ·. ; .. iw, ',q; ','.• OH

The followi:z:.greaction (XXII) represents the degradation of cyanidin• XXII.

OH OOOH~ H .. KOH -11- HO ...... OH ~

Cyanidin Phloroglucinol Protocatechnic acid Precise nature .of' the phenyl residue in position 2 and the points o'£ linkage of the sugar residues have been 14, •. pl'Oven by ttar:rer (6, 7, 8) and C"onsidered very reliable. Later degPadations of· sugar :free pigm.ents with dilute barium or sodium hydroxide · (10%J in au at:roosphere of hydro­ gen; also degradation~ :with hydrogen peroxide helped to establish positions ·Of the methoxy groups of peonidin

(XIII), malvi din {XIV), and hirsutidin (XV). The hydrogen peroxide used by Karrer opened the ring structure between the 2 and 3 oar'bon atoms wi. thout removing the sugar i"'esi!"'" dues ·O"P the methoxy groups. Karrer also found the .position of the sugar on the anthocya.nins by methylation of the antho• cyanins followed by the removal or the sugar group and the identification of the unmethylated position that the sugar had originally occupied. This la.tteJ;l' method estab1ished the location of the sugar residue o:f the monoglucosides as always being in the 3 hydroxyl position of the anthocyanin nucleus. In the. ease of the diglucosides he found that the second sugar residue i.s gene:raJ.ly, but not always, attached to the 5 position. The procedUI"e .followed by most investigators today is the degradation of the anthoeyanidin with 10~ barium hydroxide in an inert atmosphere in order to prevent side oxidations. The earliest conclusive contributions in the syn... thesis of anthocyanidi.ns and the anthocyanins ware made by .1.5~.

Wlllstatter (22, 23, 24, 25 11 26) and the Robinsons {10::~ 121

1.3, 14 1 1.5) worki!l..g independently. ·. rii11statter, the first investigator to synthesize anthocyanidin• used Grignard · reagents with ,3 methoxy eourflB.:rl.ns to produce the desired anthocye.nidins. He used thh1 method to prepare pelargoni­ din and cyanidin. This general metliod 1 the addition of ·• aryl ·Grigns.rd reagents to cou.marins, is ·shown in the follow­ ing equation (XX.IIJ: }.,

Willstatters sYl3-thesis of pelargonidin is: p-an.isylmagnesium bromide {XXIV} reacted with 3, $, 7 tr.imethoxyeoumarin (XXV) which. produces the carbinol base. pelargonidin tetramethyl ether (XXV!) which is acidified to the chloride of pelar­ gonidin tetramethyl ether (XXVII)~ The ether is then heated with concentrated hydrochloric acid in a sealed tube causing the m.ethoxy gz-oups to hydrolize off to give pelargon1din chloride. XXIV. xxv. 16. T.lVI•

Robinson on the otlc~Bl"' hand emplC?yed a second geners.l method, the condensation of o-hydrcJt:ybenzaldehydes -vdth eppropriate ketones to forr11 2 h.ydroxyehalcone v.;hich is con­ verted to the anthocya.nidin vd th hydrochloric scid causing ring closure. This method is shown by the following equa- tion (XXVIII). XXVIII.

OH

CHO

X

Robinson, who pN\bably did more work on synthesis of antho­ eye.nidins tb.an any other investigator used this second method with success in the synthesis of all the anthocy­ anidin types. He even synthesized several of the natural .occurring snthocyanins, a more difficult and much more importe,nt achievement, since his method proved tha. t antho- .. cysxdns could be made in the manner made in nature instead of being in the forrn of a salt~ Th.ese investigations of Robinson, as tr:ell as Willstatter, have confirmed the parent types of. the elass, originally assigned by Willstatter, namely pelargonidin, cyanidin end delphinidin.

Other early inveHtigRtions ~~ere s.tter:1pted by Ever- est (3, 4) who reduced flavones w:i th zinc dust and hydro• chloric s:cid; V.Tillstatter and Ne.llison (23) who reduced in methyl alcoholic hydrochloric acid to cyani­ din chloride using magnesium$ zine dust or sodium amalgam as a source of hydrogen; Malkin and Nierenstein (9); Asahina ru'ld Inubuse {1) and others.. None of these investi• gations, ho-v1ever, were as significa.nt as those of' tvillstat­ ter and Robinson. Some were merely qualitative, others produced exceedingly small yields~ still others gave prod­ ucts which were diffieult to isolate. Since the members of this group are found in the sap of plant cells, . they s,re all to some degree soluble in water, quite soluble in hydro.xylic solvents and insol- uble in. non-hydroxylic solvents such as ether, benzene and chloroform. Characteristic color res.c tions ltrere investige,ted early by the Robinsons (16, 17, 18, 19). These qualita- ! ;

18. . . tive methods aile considered very aecurate andare useable for crude extracts since only one pigment is usually iD,.;.: vol.ved in the color of a plant;. Some of these importa~t properties of the anth()eyanidins are summarized iri the following paragraphs: . . Delphinidin is blui-sh. red in aqueous solution~ pel- argonidin red while the other four are vio'let red.. In water pelargonidin, delphinidin and peonidin are readily·· . . soluble while the others are only slightly soluble.

Cyanidin and delphinidin will give amyl a~eohol an in­ tense blue color on the addition of sodium. acetate and ferric chloride, while the others give no reaction to this test. Fehlings solution can be reduced in the cold by cyanidin and delphinidin whereas pelargonidin must be warmed and the other three anthoeyanidins :must be boiled to produee the same r-eduction. In soda solution aJ.l types give a violet to blue color change. In aqueous so1ution the red color of' pelargonidin fades on standing, that of' delphinidin fades slowly in the cold• rapidly when heated. 'l".ae other anthocyanidins must be heated for the eolo:r to .fade. Addition of 10?.{ aqueous sodium hydroxide to a dilute solution of an tho cyanidin which i s then shaken with air will destroy petunidin and delphinidin at once while other 19. anthocyanidins are relatively stable to this test. All can be extracted to some. extent by an organic mixture of anisole and ethyl isoamyl ether w1 th .5% :pi erie acid ex­ cept delphinidin. A pink tint is imparted to a mixture o.f eyelohexanol and toluene by cyanidin while malvidin gives the mixture a .faint blue color. Peonidin and_ pelar­ gonidin are extracted in large amounts while delphinidin and petunidin are not extracted at all. Distribution between 1% hydrochloric acid and a mix­ ture o.f eyclohexanol (1 volume) and toluene (5 volumes) indicates that delphinidin and petunidin are not extracted. Mal vidin imparts a faint blue color to the organic layer­ cyanidin a pale rose color. Peonidin and especially pel­ argonidin are extracted to a considerable extent. The Robinsons (18) concluded that certain substances called ~o-pig.111eD:ts influenced color anthocyanin and sntho­ eyanidin solutions. These .e:f..fects e.re detected in strongly acid solutions and the presence or absence of these sub­ stances is undoubtedly a. factor to be taken into eonside:ra­ tion, These natural oceu:rring eo-pigments include anthox­ anthins (.flavones and flavonol saccharides, e te.}, tannins and other substances. Circumstances point to eo-pigmentation of anthoeyanins, for example the pelargonin salt in the flower -. Finally probably traces of iron and other 20. inorganic substances may affect flower color. As an ex­ ample 1 t has been observed that when stalks of red hydran­ gea are immersed in very dilute aqueous ferric chloride the flowers_ s_lowly become blue. Ashes of many flowers contain 1 to 2% ferric oxide and the a.nthocysnin test for iron is one of the most delicate knoll>n. As mentioned earlier these pigments are amphoteric substances whieh form true oxonium salts with acids.

1"hese salts are remarkably stable and have extraordinary crystallizing properties. Consequently in final stages of isolation the pigment is usually converted into its hydrochloric or picric acid salt·• Acid salts of antho ... cyanins and anthocyanidins are usually red, metallic salts with bases blue and neutral salts purple. Anthocyanins have an affinity for 1netals as noted in the canning industry by deteriorization of the tin cans and the color change or the canned .fruit. This is due to the fact that metal salts of the anthoeye.nins are insoluble and therefore precipitated on the inside of the can.

Strong absorption powers are possessed by the antho­ cyanins and anthocyanidins over a wave length range of 6000 to 2000 angstrom units. T'.ne sugar free. pigments and

the glycosides are practically the same. r.faximum absorp­ tion, the cause of color, lies in the visible spectrum with all types at approximately 5000 angstroms and all have a band in the vicinity of 2700 angstrom$.

Flowers and fruits of plants contain many dif~erent glycosidic combinations o:f various anthoeyanins, with vary­ ing attachments of sugar molecules yielding different com­ pounds of varying color and shades of colo:r, The amount o:f anthoeyanins in flowers snd plants val'ies over a v.ride range.

Typical extremes are cyanin~ only 0. 75% dry petal weight in blue corn flower and 14% in the dark green garden variety, and in red dahlias over 20.0%. Anthoeye.nins in addition to being f'ound in flowers and f'rui ts are a.lso present in leaves of' eei>tain trees. Isolation of the principal types usually proceeds along the following lines. Extraction is earried out with methyl or ethyl alcohol .followed by hydrochloric acid which is bubbled through the extract to change the anthocyanin to its corresponding anthocyenidin. This crude chloride is then proeeipi tated by the addition of ether, purified by redissolving in aqueous hydrochloric acid, followed by add­ ing quantities of alcohol and ether foro reprecipite.ting the salt. Final recrystallization can be accomplished with alcoholic hydrochloric acid or aqueous alcoholic hydro­ chloric acid. Recent investiga.to!l?s have .found that some anthocy­ anins e~~ be purified through chromatogr-aphic absorption. Much work has been done in this rie1d and many articles have been publiShed reporting the progress made. · EXPERIMENTAL RESEARCH

Petals of Canna InQ.ic.a :flowers were picked !'rom pls.nts in the southeast eorner of the College of the Pacific campus. The Pure red colored ·petals were carefully separated. for thi.s research, and the yellow and mul tieolored petals discarded. The petals were then dried in the open air for two weeks. ~Yhen thoroughly ··.di'>y. they were mixed with dry meth­ anol in a Warring blender.. Dry hydrogen chloride gas was generated b'y dJ:oopping concentrated hydrochloric acid into concentrated sulfuric acid and bubbled into the methanol petal mixture until saturation. Oolor o£ the mixture was dull brownish red at the beginning of this step and a bright red color at the end indicating the formation of an antho­ cyanidin. The methanol containing the anthoeyanidin was then filtered out using aspirator filtration and the extracted petals wei"e discarded as ws.srte.

The a.ddi tion ot ~Y ether to the methanol and antho­ eyanidin caused the formation of: a somewhat dUll reddish jel like precipitate. Excess methanol was driven off by heating leaving a supersaturated solution or anthoeyanidin. Evaporation was ea:rried out under vacuum slightly over. room temperature to avoid decomposition. 2).

~he supersaturated methanol solution was then cooled and the crystallized anthoeyanidin separated out by .filtration. The crystals were turther dried in a vacuum dessicator to constant weight. Color of the ceystals was reddish brown. The crystals were tested for solubility in water and various other solvents. '.rh.ey were found to be some ... what soluble in water giving a light red colored solution which color faded on standing. The crystals were moderately soluble in methanol md ethanol, insoluble in ether, benzene, carbon tetrachloride and other non polar solvents. They were only slightly soluble in the "universal solvent" di­ methyl formamide. In ten percent sodium hydroxide solution color changed to brownish blue indicating formation ot the color base salt, the blue anion, as explained on page 6, and typical of almost all the anthocya.nidins. A second pigment extraction was carried out with amyl alcohol giving a blue green extract. The color deep ... ened to an intense blue upon the addition or dry hydrogen chloride gas. This extract was filtered and dry ether added to the filtrate. The excess amyl alcohol was driven out by heat, the supersaturated solution cooled, filtered and the cry·stals dried. The addition of sodium acetate and a trace of ferric chloride produced a color change somewhat to blue green 24. and on heating to a dark blue bla¢k4 Cold Fehlings solu­ tion caused the •. eolor to change to blaek and on heating sligptly to a ,reddish prawn showing the reduction of the Fehlings solution, ·Sodium eal'bonate. solution-of. the erystals·was.reddish bro"Wn in color turning to a blue wine color "Wh~n warmedi-. A third extraction. made with a mixture of one vol­ ume of , eyelohexanol and 5 volumes of toluene extracted· enough pigment to color the .solution a pinkish orange .• Dilute bal>ium hydroxide vas added to the solution until a da:z>k eoloxaed precipitate settled out. A decompo­ sition point could not be dete12mined because of the dark color. The addition of a small amount o:f hydrogen ehlo12ide caused the formation of the anthoeyanidin again as a pre­ cipitate, assumed to be the barium sel.t of canna indica snthoeyani.din. A weighed amount of dry petals was placed in a so:xlet extraeter and. methanol used to carry out another extraction. The resulting extraction was cherry red in eolor, 1rrhich. color deepened to darker red on the addition o:f hydrogen chloride. _ Passing ether into the mixture again resulted in the formation of dull reddish colored jel like precipitate. Again the excess methanol was driven o.ff by 25. heat and the supersatUZ.ated solution allowed to cool~ pre ... cipitating the anthocy~~din• which was then -ated. by filtration. • Afte~. dryi~g these crystals to coristant weight in a vacuum 'dessicato:tt the mean percentage of antho­ cyanin in the dry petals was ealeUiated to be zi.aa 'percent • . 1'h~ apectrUm. oi' ihis canna indica anihoeysnidin in ethanol. was observed in a Beekman Model :S spectrophotometer :from .)60 to 650 mi11imierohs •. The eanna anthocyanidin was ·fused with potassium h7"" droxide, yiel.ding as chief degradation pxooduets a phenol. and a potassium sal. t of a phenolearbo:qlie aeid. Diethyl ethexo was added to the alkaline solution and the ·ethel" evaporated• The remaining residue melted at 218°0 ... 220°C, as did a mixt~e of the residue end phlot~oglueinol. Phloroglucinol, anb.Jdrous, melts at 2l9°C. The degradation products were heated with methanol and ,hydrogen chloride gas to give a methyl este:r of the phenolcarboxylie aeid. This methyl ester melted at 1290 c, the melting point of methyl hydl'oxybenzoa te which has been reported as having a melting point of 127°0 - 129°C and 131°0. These products recovered from the alkali .fusion are

. . phloroglucinol and p-hydroxybenzoic acid which would indi· eate that the canna anthoeyanidin was composed of these units. The strue ture of the anthoeyanidin containing these 26~ unit.s w"'uld be 3~. S~ 7,.. 4• tetrahydrtoxy 2 phenylbenzop:ypil• ium <>r pela.Pgomdin {.XXIX}.. :, xxrx •.

HO OH

OH

The pigment of Canna Indica pure red colored petals was extr~cte'd w1 th methanol and precipitated .as a chloride. Color was ~i.ght red fn the ac.id. solution indicating fo:rma­ tion of an. anthocyanldin. The. anthocyanidin was crystal,;. lized out of the. solution and the crystals tested for solu­ bility. · They· we're moderately .. so 1 ubl e in .methanol and

' . . ethanol, somewhat soluble in water, imparting a light red color to the solution, which color .faded on standing. They were only slightly soluble in dimethyl formamide and insol­ uble in nonpolar organic solvents. The pigment colored 10% sodium hydroxide brownish blue, gave a pink color to a mix­

ture of eyclohexanol and toluene and reduced Fehlings solu... tion when warmed. Degradation products of the pigment were phloroglu­ cinol and p-hydroxybenzoie acid as proven by melting points ..

.A spectrum was obtained of the pigment in the range of 360 to 650 millimicrons with a Beckman spectrophotometer. Coloring matter in the Canna Indica flowers was thus shown to be an anthocyanin since it was red in acid solution and bl.ue in alkali. Melting points of the degradation products show the pigment to be pelargonin. BIBLIOGRAPHY

1. Asahina, s. and Inubuse, J., ~~~ 61B, 1696 {1928).

Cram, D. J. and Hammond, G. s., Organi~ Chemistry, McGraw-Hill Book Co., Inc., New York (1959}.

Everest, M., ~· Roy. • .§2.2.• London (B), 87, 444 (191!+).

Everest, }f.,~· Roz. §..22_. London (B), 88, 326 (1914).

Gilman, H., Organic Chemistry, ~Advanced Tres.tise, John Wiley and Sons, Inc., liew York (1953).

6. Karrer, P., Helv. £h!E!• ~., 10, 67, 729 {1927).

Karrer, P., ~· ~· ~., 12, 292 (1929). 8. Karrer, P., Helv. Chim. Acta., 15, 507 (1932}. 9. Malkin, A. end lUerenstein, R., !!!:•, 61B, 791 (1928).

10. Robinson, F. R. s • ., Nature, 137, ·~4-95, 172-173 (1936).

11. Robinson, G. 11., l.• ~·Soc., 1939, 1605, 1606, 1607 (1938). Robinson, J. Soc. Chem. Ind., 12. R., ...... ,_.. ~ __..,_. 52, 737 (1933}. 13. Robinson, ~., Nature, 132, 625 {1933). 14· Robinson, R., ~., 67A, 85 (1934) • 15. Robinson, R., Nature {Royal Jubilee ~:rumber), 135, 732 (1935).

16. Robinson, G. M. ~d Robinson, R., Biochem, l• 1 25, 1687 (1931). 11. Robinson, G. M.. and Robinson, R., Biochem. l_., 26, 1647 (1932). 18. Robinson, G,. r-1:. and Rol:>inson, R,., Biochem. J., 27, 206 (1933). - ' 19. Robinson, G. M. and Robinson, R., Biocbem. !!.q 32, 1661 (1938).

20. Rupe, H., ~ Ghemie ~ Naturlichen Farbstoffe$ Vieweg

and Sohn, Braunschwei~ (1900).

21. St. Jonesoo, Co:mpt.~., 174, 1635-1637 {1922). JO. 22,. Wi,ll'sta,tter., R .. , Ann., 40l, 189 (l913h ~-.,

2.3,. Will.stRtte:t', R~~ ~ .. , 47,· 2865. (1914) •.

147 (1915Y.,

25 .• Wills.tatter, .R.~··.ABB·•·4l:a, llJ,. 1.37, ·149, 164. 178, 195: 217, 231 (1916).

26. Wills tatter, R., ~H 57, 19.38, 1945 '(1924).